Factor Analysis Lecture 1. Flashcards
What is Factor Analysis
1: Variation: A series of statistical methods to understand the variance of how people answer items and perform on tasks.
It accounts for variation. It account for if each variable is assessing a different thing.
It reduces a lot of variables to a fewer number of variables so its easier to digest.
2: It gives us a way to see from the data what model we should use.
What are the uses of Factor Analysis?
1: Understanding:To understand the structure of a set of variables.(eg Spearman used Factor analysis to understand IQ)
2: Construction:To construct a Questionnaire to measure an underlying variable (eg questionnaire to measure burnout).
3: Data reduction: To reduce a data set to a more manageable size while retaining as much of original information as possible.
4: Determine: Helps us determine if all items measure different things or if they reflect the same construct?
What is Factor analysis driven by?
FA is data driven. It gives us a way to see from the data what model we should use. The data strongly influences how we come to think about what when we ask people questions/observe people/give them tasks.
At it’s heart what does FA look at?
Correlations. They are at the core of understanding relationships between variables. FA examines the patterns in data to see which explanation is supported. It takes raw data and looks for patterns.
What is a Correlation Matrix?
A correlation matrix is a table which allows you to see which items are related to each other. It gives us numbers that tell us the strength of relationships between variables. The higher the number, the higher the relationship.
A table of correlation coefficients between variables.
What is a Latent Variable?
A variable that cannot be directly measured. It hints at a grouping variable that brings all these disparate variables together according to a common factor.
Eg: You can’t measure burnout directly but you can measure different aspects of it like motivation and stress levels.
What is Pearson’s correlation?
The strength of relationship between two variables. How much X varies from Y. Denoted by R or r. Goes between -1 and +1. A coefficient of +1 indicates two variables are positively correlated.
What number are Correlations typically at?
They are rarely greater than .5 (>.5) Usually be between .2 and .3
What are the controversies surrounding Factor Analysis.
Oversimplified use and interpretation.
SPSS will always produce a result.
